"""
Baseline models contains models using default model parameters

"""
from sklearn.svm import LinearSVR
from sklearn.tree import DecisionTreeRegressor
from sklearn.ensemble import (
    BaggingRegressor,
    AdaBoostRegressor,
    GradientBoostingRegressor,
    HistGradientBoostingRegressor
)
from xgboost.sklearn import XGBModel
from sklearn.linear_model import LinearRegression
from sklearn.pipeline import Pipeline


""" Linear Regressor """


def linear_regressor_pipeline_factory(model_params={}):
    """Linear Regression pipeline factor"""

    linear_regressor = LinearRegression(**model_params)

    return Pipeline(steps=[("linear_regressor", linear_regressor)])


""" SVM Regressors """


def svm_regressor_pipeline_factory(model_params={}):
    """SVM Regressor Pipeline"""

    linear_svr = LinearSVR(**model_params)

    return Pipeline(steps=[("linear_svr", linear_svr)])


def svm_bagging_regressor_pipeline_factory(model_params={}):
    """Ensemble of Bagging Regressors"""
    linear_svr = LinearSVR(**model_params)
    bagging_ensemble = BaggingRegressor(base_estimator=linear_svr)

    return Pipeline(steps=[("bagging_ensemble_of_linear_svr", bagging_ensemble)])


def svm_boosting_regressor_pipeline_factory(model_params={}):
    """Ensemble of Bagging Regressors"""
    linear_svr = LinearSVR(**model_params)
    boosting_ensemble = AdaBoostRegressor(base_estimator=linear_svr)

    return Pipeline(steps=[("adaboost_ensemble_of_linear_svr", boosting_ensemble)])


""" DecisionTree Based Regressors """


def dct_regressor_pipeline_factory(model_params={}):
    """SVM Regressor Pipeline"""

    dct_regressor = DecisionTreeRegressor(**model_params)

    return Pipeline(steps=[("dct_regressor", dct_regressor)])


def dct_bagging_regressor_pipeline_factory(model_params={}):
    """Ensemble of Bagging Regressors"""

    dct_regressor = DecisionTreeRegressor(**model_params)
    bagging_ensemble = BaggingRegressor(base_estimator=dct_regressor)

    return Pipeline(steps=[("bagging_ensemble_of_dct_regressors", bagging_ensemble)])


def dct_boosting_regressor_pipeline_factory(model_params={}):
    """Ensemble of Bagging Regressors"""
    dct_regressor = DecisionTreeRegressor(**model_params)
    bagging_ensemble = AdaBoostRegressor(base_estimator=dct_regressor)

    return Pipeline(steps=[("boosting_ensemble_of_dct_regressors", bagging_ensemble)])


def gbt_regressor_pipeline_factory(model_params={}):
    """Ensemble of Gradient Boosted Trees"""

    gbt_regressor = GradientBoostingRegressor(**model_params)

    return Pipeline(steps=[("gbt_regressor", gbt_regressor)])


def xgb_regressor_pipeline_factory(model_params={}):
    """Ensemble of Gradient Boosted Trees"""

    xgb_regressor = XGBModel(**model_params)

    return Pipeline(steps=[("xgb_regressor", xgb_regressor)])


def histgbm_regressor_pipeline_factory(model_params={}):
    """ Histogram GBM - This is very similar to LightGBM """

    hist_gbm = HistGradientBoostingRegressor(**model_params)

    return Pipeline(steps=[("hist_gbm", hist_gbm)])